Simplify management of ecryptfs protected directories to be able to:

-Easily mount an encrypted directory in recovery mode
-Easily transport an encrypted directory easily (like saving/mounting the encrypted dir to a single image file)
-Have the ability to tell nautilus to mount the encrypted directory with "archive mounter" and creatively find a simple way to decrypt it.
-Have the ability to graphically manage keys and passphrases

Sorry I can't get an a better title than that, but I'll explain what I'm talking about.

The Desktop's performance is a huge figure in the feel. We all know this.
Now the issue we have here is that Ubuntu's default desktop is just far to separated in its code base. The default code base of Ubuntu should be unified it should as be fast. As of now what we have is horribly inefficient and on a path to get worse from what I've seen with project surrounding. If you want to see for yourself just use a bare Xfce 4.6 desktop no Ubuntu/Python/Mono tools and compare it against Ubuntu/GNOME's performance.

Now there are some heavy advantages we can get with code unification. One is that code can easily be shared between API's, without heavy use of "gluing" things together like what we do with Python and such. Sharing code is what open source is about and we should not be making this harder.

We also want speed. GTK+ made a bold and strong choice by building around C. C is blazingly fast, and C is also the most portable language. Sadly C's design is so general and low level that programmers only want it for drivers, because drivers designed in C are normally the best. Thankfully the GTK+ crowd took another strong step and designed the GObject library, an object like C library for GTK, from which we have easier languages. Like Vala, a C# syntax language, and Genie, a Python syntax language, which are built on C/GTK+ because they in fact mid compile become clean C/GObject/GTK+ code, with comments placed correct might I add, meaning no glue is needed.

So for what we have available this task isn't as hard as one might think. Though just because it's easier now than its been before we shouldn't wait around and we should begin immediately. Beginning with the heavy weights Python/Mono, C++ on the other hand I know isn't something to mess with, example getting Webkit ported from C++ to C is an insane idea on another note we can always have an alternative basic C built html viewer made if a system cant have C++.


Quick note summary :
Advantages to the default install:
More portability. ( Some know the pluses in this. )

[....]

Ubuntu Netbook Remix (UNR) came along with some netbooks, but most of these netbook users have not ever used a linux based OS. Some terminology such as Add/Remove... and The Terminal, sudo: , packages, repositories, etc.. made really hard for them to use URN. In consequence URN has lost a very important opportunity to enter establish himself as a more popular OS. When UNR fisrt came out, it was came pre-installed in almost 13% of the netbooks, now a days this number has reduced to near 5%.

(When I talk about "the netbook users", i mean the people who bought a netbook with UNR, but were not really familiar with it)

Ubuntu offers us a very secure, highly configurable environment, but when we move to the netbooks the situation this is really different. The netbook users want a OS which is fast, stable and easy to use. The terminal; for example; for most of us is really helpful as it let as do all kind of tasks. But if someone who has never used this OS try to use it, he will don't know how to use(at least not completely well). This is the first issue, in order to the UNR to be widely accepted, int needs to be much easier to use. Some users has already posted similar ideas, as the use of universal package format. Another important problem is that the netbook users does not know the Ubuntu strengths, as the security against intruders or viruses, in which we are far superior to Windows.
With UNR 9.04 much of the netbook users did not figured out how to use it correctly and most of them were seemed forced to return their netbook or change it for another one with Windows. That is why we need to make UNR much easier.

I have it physically disabled on my desktop but the beep on my g/f's work laptop would put most sonar's to shame! It doesn't have a bios option to disable it and I don't think her boss would appreciate me taking a screwdriver to it either so, every time tab won't complete a line or backspace doesn't have a character to delete we're subjected to a this stupid 100 dB sonic weapon.

Naturally this is even more of a problem at night or in the library.

Now I know it can be disabled by blacklisting and removing a kernel module which is what I now do BUT... I'm certain this damned beep annoys more people than it helps and as such I think something ought to be done about it...

Because netbooks are machines that are made for a very specific task that is quite different from 'normal' desktop use, there are several programs installed that 95% or even 100% of the users will not use. For example:
- Brasero. I'm pretty sure that almost nobody uses his netbook to burn cd's.
- XSane Image Scanner.
- Assistive Technologies. The size of the devices doesnt make it very accessable for the less capable.
- PalmOS Tools. Who does even use a palm these days anyway?

hi,
I am using linux past four years by the time i was using various distros. Most of linux operating systems are huge and CPU&RAM hungry.

Most of the people using ubuntu for Internet purpose only.So make a new derivative(like Kubuntu,Xubuntu) and it should be small in size less than 50 MB. IT can be able run in 400 MHZ processor,32 MB RAM(minimum requirement).and better internet connection..

---- All Apps must have GUI interfaces only --

so we can able run in netbook, Mini ITX based PCs... also people who need linux for internet only..

I hope canonical should release an Official version of Ubunti derivative with unlimited support. Because Slitaz, DSL, etc. comes without a good support for end users.

Many times when testing ubuntu in alpha-stages there can be problems with booting/loading the live-cd. I tis difficult for most people to document what is going wrong during boot-up of the live-cd. How do one save and retrieve such data? The big problem is to log and document every bug or error when using the live-cd and it stops before it loads.

Would be very convenient to have more ready localized installation live-iso-media, but to be able to fit it in 4 gb (or even 2 gb) usb-memory-stick.
dvd-iso is oversized for that and cd-isos don't include all language packs, so they don't offer localization-ready usb-memory live-mediaset.

So the main thing that I suggest is to offer more fully localized installation ISOs. Now there's plenty of unnecessary locales in dvd and not adequate in cd.
Of course I and some other geeks know how to make a more convenient installation media by adding packages and removing unnecessary, but why to force thousands to do that when it could be done more efficiently by offering some more localized installation ISOs.
The ability to run Ubuntu from usb-memory stick is a great advantage compared to Windows, so why not take the full advantage and ease out of it and to offer even better, more fully localized installation ISOs.

For example here in Finland (where linux is from btw) with Ubuntu cdISO I have to add missing localization packages to get fully finnish Ubuntu-live-usb-media. Or if I use dvdISO I have to rip off extras to give more free space in live-usb-memorystick to use that space for store space while using live-usb-Ubuntu. So much extra job with every Ubuntu release.

It's wonderful that there's Ubuntu for Gnome, Kubuntu for KDE, and Xubuntu for XFCE, and they all do well in supporting their own Desktop Environment, and I won't say it's ridiculous to have them separated the way they are. We do know, however, that all three, including all the 'boxes and FVWM and Enlightenment play very well together in a Debian environment together.

It's also very good that the LiveCD install for Ubuntu is very neat, straightforward, and doesn't offer more options than new users know what to do with.

I've also seen propositions for choosing what packages to install, but I don't think picking out every little package is quite the right idea.

It could be to the benefit for many if there were another 'distribution' put together meant for installs where broadband was readily accessible, one that isn't meant for a streamlined installation of one DE, but instead gave you a simple and descriptive way of choosing DE's, one or many, or even not choosing one at all if someone just wanted pseudoterminals.

Then the user could choose their login manager, if applicable, and be informed the differences between, well, basically GDM and KDM. I don't know of any other login managers that really give users what they need.

The 'roles' of the computer could also be selected since computers have so many different situations.

After the choices have been made, the necessary packages, already up to date, could be downloaded and installed.

This wouldn't interfere with X/K/Ubuntu and give interested users more choices installing Ubuntu without the painful process of sifting through many different applications.

What's even better is this installer could be updated every time the LiveCD is started since it would essentially be expected to. This would make changes in your options independent to the image you download and burn. In fact, you could never need to download another install CD again, the installer would find out every time what it has to do to install the next distribution from files stored on a server.

[....]

It´s not easy to switch OS to the experienced users. And for most simple men it´s unbelievable.